I started a PhD nearly four years ago. At that time, my supervisor provided me a project, but after more than one year we had to change it: the instrumention was to old and he didn't want to spend monay on it.
We did many other attempts to figure out something else to do. At the end of the second year I did some reasonable measurements that could provide some interesting information, from that time I realized what could be my PhD project!
Therefore, after spending 2 years on useless things, in four mounts my supervisor wrote and published a paper and I am the first author! Today I am close to the viva, and I just realized that the paper he wrote is totally wrong!
I don't want to waste my career because of a bad PhD management. The question is: can I write a PhD thesis based on my personal considerations about that measurements and completely ignoring arguments writen on a published paper?

The thesis can be separate from the published paper, but your advisor is going to read it so . But are you sure the published paper is totally wrong?

I confirm that the paper is wrong. I realized it when I started writing the introduction. In that case I had to collect all the information I read in articles and books and I found a large amount of evidences of its mistakes. I personally believe that the paper is not only wrong, but there are also false proof of the final results.
The best thing I could do would be writing a chapter results based on my personal considerations and trying to not contradict too much that paper, which is in some percentage contradictory by itself. It's not an easy job, but it could be safer and scientifically more correct than just writing down that paper on my thesis.

I am sure that the paper is not only wrong, but there are also some fake results. My PI knows that I can hardly defend that work, for this reason, he suggested me to just referee that paper about some topics that can not be discussed to much